Mobile rescue intelligent management method based on RFID automatic identification and AI artificial intelligence interaction

Abstract

The invention discloses a mobile rescue intelligent management method based on RFID automatic identification and AI artificial intelligence interaction, and belongs to the field of mobile rescue intelligent management. According to the invention, the problem of low efficiency of medicine and material inventory management of the rescue carriage in an inpatient area due to an existing manual inventory mode is solved, batch accurate identification can be carried out by adopting an RFID remote identification technology, batch inventory is realized to replace one-by-one checking, the inventory efficiency is effectively improved, and the inventory management cost is reduced. Manual electrocardiogram data checking is replaced by image visual recognition of electrocardiogram data, secondary recording is not needed, rescue data can be automatically recorded, rescue efficiency is improved, medical advice recognition is assisted by AI voice instead of manual recording of medical advice, the workload of nurses is reduced, the workload of secondary recording is reduced, and rescue efficiency is improved. And through a rescue scene full-process digital AI closed loop, digital tracing can be completed in the rescue process, and the problems that traditional handwritten records and data are not real-time and inaccurate are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile ambulance intelligent management, and specifically to a mobile ambulance intelligent management method that interacts with RFID automatic identification and AI artificial intelligence. Background Art

[0002] Emergency vehicles are essential equipment for hospitals, specifically designed to store emergency supplies, medicines, and equipment. The quality of medication and item management within emergency vehicles directly impacts the success rate of patient rescue. To ensure the smooth progress of rescue operations, emergency supplies within the vehicle must be of guaranteed quality, quantity, functionality, and integrity to fully safeguard patient care. However, nurses must individually check and verify the contents of the emergency vehicle each day during shift changes. The diverse range of medications, inconsistent quantities, and inconsistent expiration dates within the vehicle make verification and management difficult. Therefore, this approach does not meet existing needs. Therefore, we propose a mobile emergency management method that integrates RFID automatic identification with AI artificial intelligence. Summary of the Invention

[0003] The purpose of the present invention is to provide a mobile rescue intelligent management method that interacts with RFID automatic identification and AI artificial intelligence. By adopting RFID long-distance identification technology, batch accurate identification can be performed, batch inventory can be realized instead of checking one by one, and electrocardiogram data can be recognized by image vision instead of manual review of electrocardiogram data, thereby improving efficiency. AI voice-assisted medical order recognition is used instead of manual handwriting of medical orders, which reduces the workload of nurses and the workload of secondary entry. Through the digital AI closed loop of the entire rescue scene, the rescue process can be digitally traced, solving the problems raised in the above background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a mobile emergency intelligent management method that integrates RFID automatic identification and AI artificial intelligence interaction, comprising the following steps:

[0005] Step 1: Use the smart portable terminal to create a database of all medicine information for each emergency vehicle, including quantity and category, and attach RFID electronic tags to each medicine and consumable;

[0006] Step 2: Based on the preset coding rules, unique identification codes are assigned to the materials and medicines in the emergency vehicle. When initializing the materials in the emergency vehicle, the electronic tags are scanned by a portable RFID device, the drug name and expiration date information are associated and synchronized to the hospital's emergency vehicle material management platform;

[0007] Step 3: The RFID portable device uses its internal radio frequency module to send microwave field strength to activate the drug's RFID electronic tag. The electronic tag information is actively uploaded to the RFID portable device and decoded and translated into the drug material number through internal modulation and demodulation and APP program decoding. The portable terminal system then wirelessly communicates with the back-end hospital emergency vehicle material management platform and obtains the product name, material information, and expiration date corresponding to the drug material code by querying the database.

[0008] Step 4: The hospital emergency vehicle material management platform records the expiration date registered when each drug RFID electronic tag is initialized, and sets different near-expiration time limits based on different drug names and specifications. When a drug enters the near-expiration range, an audible and visual alarm will sound to remind the user to take action, and the alarm will be turned off after the action is taken;

[0009] Step 5: The AI ​​voice recognition system converts the doctor's medication order into medication information in real time and stores it in the portable device system. When the nurse takes the medication, the AI ​​voice automatically announces the name of the medication and the method of use, and automatically records the time the medication was taken out of the warehouse;

[0010] Step 6: Use a portable terminal to capture an electrocardiogram image and automatically extract the vital signs and physical data of the relevant patient;

[0011] Step 7: Integrate all medication information during the rescue process, including product name, quantity, medication time, electrocardiogram changes, and whether the medical order is closed-loop, to form a rescue medical order execution form;

[0012] Step 8: Use the mobile app and the hospital emergency vehicle material management platform to scan the medicine to obtain detailed information, and query and count the emergency vehicle's operation and maintenance status in multiple dimensions based on time, emergency events, and administrator information.

[0013] Furthermore, in the step three, all drug material information is automatically counted using a portable RFID device, and the missing drugs corresponding to the unidentified RFID electronic tags and the number of layers they are on are presented through voice broadcast and visual images.

[0014] Furthermore, the missing medicines are presented through voice broadcast and visual images. The nurses further check whether there are any missing medicines. If there are any missing medicines, new medicines are added to the warehouse, and RFID electronic tags are attached and initialized, waiting for the next inventory and delivery.

[0015] Furthermore, in step 4, the process of implementing drug aging monitoring is as follows:

[0016] Basic information entry and binding: During step 2 initialization, attach an RFID tag to the drug and bring the portable RFID device close to the RFID tag. A prompt will pop up to update the drug's expiration date and confirm the product name.

[0017] By mapping and binding the RFID tag's coding information with the item's name and expiration date, drug information is automatically uploaded to the hospital's emergency vehicle material management platform.

[0018] Set the expiration date: The hospital emergency vehicle material management platform sets the corresponding expiration date according to the name and specifications of different drugs;

[0019] Real-time monitoring and comparison: Continuously monitor the expiration date of drugs. The hospital emergency vehicle material management platform compares the remaining expiration date of the drug with the pre-set expiration date.

[0020] Early warning: When the hospital emergency vehicle material management platform detects that the expiration date of the drug material enters the set near-expiration range, it will remind the user to dispose of the expiring drugs as soon as possible through voice and interface flashing, and turn off the alarm message after the user has handled it.

[0021] Furthermore, in step five, the doctor's medication order is converted into medication information in real time through the AI ​​voice recognition system, and the following process is specifically performed:

[0022] Using dual-microphone beamforming technology combined with the RNN noise suppression model, the doctor's voice source is spatially located;

[0023] An improved end-to-end recognition model based on the open-source Whisper architecture introduces a medical-specific vocabulary enhancement module, a context-aware dosage unit error correction mechanism, and real-time voiceprint verification to ensure the legitimacy of the medical order source;

[0024] The conversion results are broadcast instantly through the TTS engine, and the doctor triggers the instruction status change through the preset command;

[0025] Regular expressions and clinical terminology are combined for analysis and automatic conversion into standardized medical instructions.

[0026] Furthermore, in step six, the electrocardiogram image is captured by a portable terminal to automatically extract the vital constitution data of the relevant patient. The specific recognition process is as follows:

[0027] Image preprocessing: Call the PyWT library wavelet denoising interface to perform denoising correction on the captured ECG image, and call the OpenCV perspective correction interface to automatically detect the vertices of the ECG image quadrilateral and map them to a standard rectangle;

[0028] Vital sign data recognition: Using the open source model convolutional neural network (CNN) and the MIT-BIH dataset, we can classify heartbeats, including N, V, L, and R arrhythmias.

[0029] At the same time, the open source model LightX3ECG was used for three-lead disease classification;

[0030] Medical rule engine: Verify the consistency between model output and physiological logic, including:

[0031] Conflict detection: If HR in the text area is 80 bpm, but the heart rate calculated by the RR interval in the waveform area is ≠ 80 ± 5%, a recheck will be prompted;

[0032] Risk value interception: If ST segment elevation is detected and QTc>500ms, the portable terminal system will automatically warn of the risk of myocardial infarction and sound an alarm.

[0033] Furthermore, in step five, the AI ​​speech recognition system has a directional noise suppression module, which uses dual-microphone beamforming technology combined with an RNN noise suppression model to spatially locate the doctor's voice source, and the software layer uses the MFCC-GMM acoustic model of the open source Kaldi framework to achieve environmental noise attenuation ≥15dB.

[0034] Furthermore, in the step 1, the ledger established is the standard quantity for each material inventory, and the RFID electronic tag is pasted in a position that does not block the key information of the drug and has anti-falling properties.

[0035] Furthermore, in step 2, the coding rules include a hierarchical combination of region, hospital, department, emergency vehicle number and drug type.

[0036] Furthermore, in step six, the extracted vital signs data include patient ID, time, lead name, heart rate value, waveform and key diagnostic parameters, wherein the key diagnostic parameters include PR interval, QRS duration and QT / QTc.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1. The present invention uses RFID long-distance identification technology to perform batch and accurate identification, solving the problem that traditional medicines use barcode identification or label identification, which can only be scanned or visually observed for expiration dates and quantity and product names. By taking inventory in batches instead of checking one by one, the inventory efficiency is effectively improved.

[0039] 2. The present invention uses a portable terminal to capture electrocardiogram images and automatically extracts the vital signs of relevant patients, thereby replacing manual review of electrocardiogram data and effectively improving efficiency. Rescue data can be automatically entered without secondary entry, thereby improving rescue efficiency.

[0040] 3. The present invention uses AI voice-assisted medical order recognition to replace manual handwritten medical order recording. Automatic voice recognition reduces the workload of nurses and the workload of secondary entry. It can also retain the doctor's medical order in voice, which is convenient for subsequent tracing of medical disputes and ensuring patient safety.

[0041] 4. By integrating all medication information during the rescue process, the present invention can realize a digital AI closed loop for the entire rescue scene, thereby enabling digital traceability of the rescue process and eliminating the problems of traditional handwritten records and inaccurate data. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of the mobile emergency intelligent management method for RFID automatic identification and AI artificial intelligence interaction of the present invention;

[0043] Figure 2 An execution diagram for the AI ​​speech recognition system of the present invention to recognize medical orders;

[0044] Figure 3 This is an execution diagram of the electrocardiogram image recognition of the present invention. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] In order to solve the existing problem of the medicines and assets in the rescue vehicle needing to be counted and checked one by one by nurses every day when the shift is changed, the rescue vehicle has many types of medicines, inconsistent quantities, inconsistent expiration dates, and it is difficult to check and manage them. In addition, it is difficult to record temporary medical orders by relying on public, which easily leads to omissions and inefficiency in recording temporary medical orders, resulting in technical problems such as safety issues of patients' medication. Figure 1-Figure 3 , this embodiment provides the following technical solutions:

[0047] A mobile emergency management method based on RFID automatic identification and AI artificial intelligence interaction includes the following steps:

[0048] Step 1: Use the smart portable terminal to create a database of all medicine information for each emergency vehicle, including quantity and category, and attach RFID electronic tags to each medicine and consumable;

[0049] The established ledger is the standard quantity for each material inventory. Medical staff can quickly check the actual quantity of medicines based on the ledger during inventory, promptly discover missing or excessive medicines, and reduce inventory errors. The RFID electronic tag is attached in a position that does not block the key information of the medicine and has anti-falling properties, reducing information misreading or time wasted searching for information due to label obstruction. At the same time, it avoids the loss or confusion of medicine information due to label falling off.

[0050] Step 2: Based on preset coding rules, unique identification codes are assigned to the materials and medicines in the emergency vehicle. When initializing the emergency vehicle materials, the electronic tags are scanned by a portable RFID device, the drug name and expiration date information are associated and synchronized to the hospital's emergency vehicle material management platform. The coding rules include a hierarchical combination of region, hospital, department, emergency vehicle number, and drug type.

[0051] Step 3: The RFID portable device uses its internal radio frequency module to send microwave field strength to activate the drug's RFID electronic tag. The electronic tag information is actively uploaded to the RFID portable device and decoded and translated into the drug material number through internal modulation and demodulation and APP program decoding. The portable terminal system then wirelessly communicates with the back-end hospital emergency vehicle material management platform and obtains the product name, material information, and expiration date corresponding to the drug material code by querying the database.

[0052] Step 4: The hospital emergency vehicle material management platform records the expiration date registered when each drug RFID electronic tag is initialized, and sets different near-expiration time limits based on different drug names and specifications. When a drug enters the near-expiration range, an audible and visual alarm will sound to remind the user to take action, and the alarm will be turned off after the action is taken;

[0053] Step 5: The AI ​​voice recognition system converts the doctor's medication order into medication information in real time and stores it in the portable device system. When the nurse takes the medication, the AI ​​voice automatically announces the name of the medication and the method of use, and automatically records the time the medication was taken out of the warehouse;

[0054] Step 6: Use a portable terminal to capture an electrocardiogram image and automatically extract the patient's vital signs and physical data. The extracted vital signs and physical data include patient ID, time, lead name, heart rate value, waveform, and key diagnostic parameters. Key diagnostic parameters include PR interval, QRS duration, and QT / QTc.

[0055] Step 7: Integrate all medication information during the rescue process, including product name, quantity, medication time, electrocardiogram changes, and whether the medical order is closed-loop, to form a rescue medical order execution form;

[0056] Step 8: Use the mobile app and the hospital emergency vehicle material management platform to scan the medicine to obtain detailed information, and query and count the emergency vehicle's operation and maintenance status in multiple dimensions based on time, emergency events, and administrator information.

[0057] The technical effects of the above technical solution are as follows: traditional drug inventory uses barcode identification or manual inspection, which requires scanning or observation one by one. The process is cumbersome and time-consuming. After using RFID technology, drug information can be accurately identified in batches, replacing one-by-one verification, thereby efficiently completing inventory. At the same time, RFID identification technology is not affected by the ambiguity of drug packaging information, and accurately obtains drug names, expiration dates, etc., preventing drug waste and misuse due to inaccurate information. Electrocardiogram data is recognized through image vision instead of manual transcription, avoiding human errors and ensuring data accuracy. AI voice-assisted medical order recognition eliminates the need for nurses to handwrite medical orders, reducing workload and avoiding secondary entry errors, improving rescue efficiency, and allowing nurses to focus on patient care. The digitalization of the entire rescue scene through AI closed-loop management can complete digital traceability of the rescue process, eliminating the traditional handwritten records and the problem of inaccurate and non-real-time data. It also provides multi-dimensional query statistics on the operation and maintenance of rescue vehicles, scientifically evaluating nurses' work performance, and improving the level of lean management.

[0058] In this implementation, step 2 is based on the preset coding rules for coding. The material coding format is shown in Table 1:

[0059]

[0060]

[0061] Table 1

[0062] In step 2, after the mapping and binding of drugs is completed, subsequent data management will be carried out based on the binding results. The specific process is as follows:

[0063] Binding information upload and storage: The portable terminal uploads the drug's unique identification code, product name, expiration date, binding time, and operating nurse ID to the hospital's emergency vehicle material management platform. The hospital's emergency vehicle material management platform establishes a three-level association database for drugs, emergency vehicles, and departments.

[0064] Real-time status synchronization mechanism: After successful initialization, the hospital emergency vehicle material management platform automatically marks the drug as "in stock";

[0065] When executing step 5, the RFID electronic tag is scanned and shipped out of the warehouse. The hospital's emergency vehicle material management platform simultaneously updates the status to "used" and records the shipping time and the associated medical order ID.

[0066] When performing inventory replenishment in step 3, the new medicines are initialized and a new "in stock" record is added to the hospital's emergency vehicle material management platform;

[0067] After the validity period warning in step 4 is triggered, the nurse manually confirms the scrapping, and the hospital's emergency vehicle material management platform updates the status to "scrapped" and generates a scrapping log;

[0068] Data consistency check: Each time the portable terminal executes step 3, it requests the standard drug list for the current ambulance from the hospital's ambulance material management platform and compares the local recognition result with the platform data.

[0069] If the RFID identification number is inconsistent with the platform record, the hospital rescue vehicle material management platform will trigger an alarm and freeze the rescue vehicle operation authority.

[0070] The technical effects of the above technical solution are: by binding information upload and storage, it ensures that the hospital emergency vehicle material management platform can obtain comprehensive and accurate drug information in real time, providing a basis for subsequent precise management; through the real-time status synchronization mechanism, the hospital emergency vehicle material management platform can dynamically update the drug status, and provide timely and accurate information for drug allocation, replenishment, scrapping, etc., to ensure the availability of emergency vehicle drugs; and through data consistency verification, the local recognition results are compared with the platform data to ensure that the drug information is accurate and consistent.

[0071] In step three, all drug material information is automatically counted using a portable RFID device, and the missing drugs and their layer numbers corresponding to unrecognized RFID tags are presented through voice broadcast and visual images.

[0072] The missing medicines are presented through voice broadcast and visual images. The nurses further check whether there are any missing medicines. If there are any missing medicines, new medicines are added to the warehouse, and RFID electronic tags are attached and initialized, waiting for the next inventory and delivery.

[0073] The technical effect of the above technical solution is: the traditional drug inventory method requires nurses to manually check the drug quantity, name and expiration date one by one, which is time-consuming and labor-intensive, and easy to miss. The use of RFID portable devices can quickly identify drug information in batches, and inventory can be completed within one minute. In addition, RFID technology can accurately identify drug electronic tags at a long distance, avoiding errors caused by negligence during manual verification, ensuring that drug information is accurate, and intuitively presenting missing drugs and their number of layers through voice broadcasts and visual images. Nurses can quickly locate and check the missing situation. After clarifying the missing drug information, nurses can promptly replenish the warehouse with new drugs and complete the pasting and initialization of electronic tags, thereby improving replenishment efficiency and ensuring the completeness of drugs in rescue vehicles.

[0074] In summary, the automated inventory process reduces the time nurses spend on inventory work, allowing nurses to focus more on core tasks such as patient care. At the same time, it monitors drug inventory in real time, promptly identifies and handles missing drugs, realizes dynamic and intelligent management, optimizes inventory levels, and reduces the risk of drug backlogs or out-of-stocks.

[0075] In step 4, the process of implementing drug aging monitoring is as follows:

[0076] Basic information entry and binding: During step 2 initialization, attach an RFID tag to the drug and bring the portable RFID device close to the RFID tag. A prompt will pop up to update the drug's expiration date and confirm the product name.

[0077] By mapping and binding the RFID tag's coding information with the item's name and expiration date, drug information is automatically uploaded to the hospital's emergency vehicle material management platform.

[0078] Set the expiration date: The hospital emergency vehicle material management platform sets the corresponding expiration date according to the name and specifications of different drugs;

[0079] Real-time monitoring and comparison: Continuously monitor the expiration date of drugs. The hospital emergency vehicle material management platform compares the remaining expiration date of the drug with the pre-set expiration date.

[0080] Early warning: When the hospital emergency vehicle material management platform detects that the expiration date of the drug material enters the set near-expiration range, it will remind the user to dispose of the expiring drugs as soon as possible through voice and interface flashing, and turn off the alarm message after the user has handled it.

[0081] The technical effect of the above technical solution is: by setting corresponding near-expiration time limits for drugs, nurses can be reminded in time to handle drugs that are nearing their expiration dates, promote the rational use of drugs, reduce drug waste due to expiration, and issue an audible and visual alarm when the drug material expiration date enters the set near-expiration range, which can effectively avoid the use of expired drugs, reduce medical risks, and thus ensure the safety of patients' medication.

[0082] In step five, the AI ​​voice recognition system converts the doctor's medication order into medication information in real time. The specific process is as follows:

[0083] Using dual-microphone beamforming technology combined with the RNN noise suppression model, the doctor's voice source is spatially located;

[0084] An improved end-to-end recognition model based on the open-source Whisper architecture introduces a medical-specific vocabulary enhancement module, a context-aware dosage unit error correction mechanism, and real-time voiceprint verification to ensure the legitimacy of the medical order source;

[0085] The conversion results are broadcast instantly through the TTS engine, and the doctor triggers the instruction status change through the preset command;

[0086] Regular expressions and clinical terminology are combined for analysis, automatically converting into standardized medical instructions;

[0087] Among them, the AI ​​speech recognition system has a directional noise suppression module, which uses dual-microphone beamforming technology combined with the RNN noise suppression model to spatially locate the doctor's voice source, and the software layer uses the MFCC-GMM acoustic model of the open source Kaldi framework to achieve environmental noise attenuation ≥15dB.

[0088] The technical effects of the above technical solution are: using dual-microphone beamforming technology and RNN noise suppression model to spatially locate the doctor's voice source, combined with the MFCC-GMM acoustic model of the Kaldi framework, the environmental noise is attenuated by ≥15dB, and the environmental noise interference is effectively filtered out, thereby ensuring that the doctor's voice instructions are clear and discernible, laying the foundation for high-accuracy speech recognition. The end-to-end recognition model improved based on the open source Whisper architecture introduces a medical-specific vocabulary enhancement module containing more than 100,000 drug / operation standard terms, as well as a context-aware dosage unit error correction mechanism, which accurately identifies and automatically corrects professional terms and easily confused dosage units in the medical field (for example, distinguishing between "milligrams" and "milliliters"), which can greatly improve the accuracy and professionalism of medical order recognition. Through real-time voiceprint verification technology, The identity of the doctor who issued the medical order can be identified to ensure that the source of the medical order is legal and compliant, prevent medical accidents caused by medical orders issued by unauthorized personnel, and ensure patient safety. The TTS engine instantly broadcasts the conversion results, and the doctor can quickly trigger the change of the instruction status through preset commands (for example, "confirm", "modify"), thereby achieving real-time feedback and efficient interaction in the medical order issuance process, reducing communication misunderstandings and time delays between medical staff, and improving rescue efficiency. The regular expression and clinical terminology set are jointly parsed to automatically convert the doctor's natural language medical order into a structured standardized medical order (for example, "intravenous injection of 10mg dexamethasone" is converted to [drug: dexamethasone][dose: 10mg][route: intravenous injection]), making the medical order information more standardized and unified, and convenient for medical staff to quickly understand and execute.

[0089] In step 6, the electrocardiogram image is captured by a portable terminal to automatically extract the vital signs of the patient. The specific recognition process is as follows:

[0090] Image preprocessing: Call the PyWT library wavelet denoising interface to perform denoising correction on the captured ECG image, and call the OpenCV perspective correction interface to automatically detect the vertices of the ECG image quadrilateral and map them to a standard rectangle;

[0091] Vital sign data recognition: Using the open source model convolutional neural network (CNN) and the MIT-BIH dataset, we can classify heartbeats, including N, V, L, and R arrhythmias.

[0092] At the same time, the open source model LightX3ECG was used for three-lead disease classification;

[0093] Medical rule engine: Verify the consistency between model output and physiological logic, including:

[0094] Conflict detection: If HR in the text area is 80 bpm, but the heart rate calculated by the RR interval in the waveform area is ≠ 80 ± 5%, a recheck will be prompted;

[0095] Risk value interception: If ST segment elevation is detected and QTc>500ms, the portable terminal system will automatically warn of the risk of myocardial infarction and sound an alarm.

[0096] The technical effects of the above technical solution are: calling the PyWT library wavelet denoising interface and OpenCV perspective correction interface can effectively remove ECG image noise and correct the shape, thereby providing clear and accurate image data for subsequent recognition, reducing recognition errors caused by image quality problems, and using open source model convolutional neural network CNN and LightX3ECG, and using the MIT-BIH data set for heart beat classification and three-lead disease classification, so as to quickly and accurately identify key vital signs data in the ECG, thereby improving diagnostic efficiency and accuracy, and verifying the consistency of model output with physiological logic through the medical rule engine, performing conflict detection and risk value interception. Conflict detection can detect inconsistencies in heart rate between the text area and the waveform area, and remind medical staff to review. Risk value interception can automatically warn of the risk of myocardial infarction, thereby effectively avoiding misdiagnosis and mistreatment due to data errors or omissions, and ensuring the medical safety of patients.

[0097] Working principle: RFID technology can be used to accurately identify drug information in batches, thereby realizing batch inventory instead of checking one by one, so as to complete the inventory efficiently. At the same time, RFID identification technology is not affected by the ambiguity of drug packaging information, and accurately obtains drug names, expiration dates, etc., to prevent drug waste and misuse due to inaccurate information. It uses image visual recognition of electrocardiogram data instead of manual transcription to avoid human errors, ensure data accuracy, and eliminate the need for secondary entry, thereby improving rescue efficiency. It uses AI voice-assisted medical order recognition instead of manual handwritten records of medical orders, which not only reduces workload, but also avoids secondary entry errors, improves rescue efficiency, and enables nurses to focus on patient care. The digital AI closed-loop management of the entire rescue scene can complete digital traceability of the rescue process, avoiding the problems of traditional handwritten records and inaccurate data. It also provides multi-dimensional query statistics on the operation and maintenance of rescue vehicles, scientifically evaluates nurses' work performance, and improves the level of lean management.

[0098] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0099] While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A mobile emergency intelligent management method that combines RFID automatic identification with AI artificial intelligence, characterized in that: The following steps are involved: Step 1: Use the smart portable terminal to create a database of all medicine information for each emergency vehicle, including quantity and category, and attach RFID electronic tags to each medicine and consumable; Step 2: Based on the preset coding rules, unique identification codes are assigned to the materials and medicines in the emergency vehicle. When initializing the materials in the emergency vehicle, the electronic tags are scanned by a portable RFID device, the drug name and expiration date information are associated and synchronized to the hospital's emergency vehicle material management platform; Step 3: The RFID portable device uses its internal radio frequency module to send microwave field strength to activate the drug's RFID electronic tag. The electronic tag information is actively uploaded to the RFID portable device and decoded and translated into the drug material number through internal modulation and demodulation and APP program decoding. The portable terminal system then wirelessly communicates with the back-end hospital emergency vehicle material management platform and obtains the product name, material information, and expiration date corresponding to the drug material code by querying the database. Step 4: The hospital emergency vehicle material management platform records the expiration date registered when each drug RFID electronic tag is initialized, and sets different near-expiration time limits based on different drug names and specifications. When a drug enters the near-expiration range, an audible and visual alarm will sound to remind the user to take action, and the alarm will be turned off after the action is taken; Step 5: The AI ​​voice recognition system converts the doctor's medication order into medication information in real time and stores it in the portable device system. When the nurse takes the medication, the AI ​​voice automatically announces the name of the medication and the method of use, and automatically records the time the medication was taken out of the warehouse; Step 6: Use a portable terminal to capture an electrocardiogram image and automatically extract the vital signs and physical data of the relevant patient; Step 7: Integrate all medication information during the rescue process, including product name, quantity, medication time, electrocardiogram changes, and whether the medical order is closed-loop, to form a rescue medical order execution form; Step 8: Use the mobile app and the hospital emergency vehicle material management platform to scan the medicine to obtain detailed information, and query and count the emergency vehicle's operation and maintenance status in multiple dimensions based on time, emergency events, and administrator information.

2. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In the step three, all drug material information is automatically counted using a portable RFID device, and the missing drugs corresponding to unrecognized RFID electronic tags and the number of layers they are on are presented through voice broadcast and visual images.

3. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 2 is characterized in that: The missing medicines are presented through voice broadcast and visual images. The nurses further check whether there are any missing medicines. If there are any missing medicines, new medicines are added to the warehouse, and RFID electronic tags are attached and initialized, waiting for the next inventory and delivery.

4. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In step 4, the process of implementing drug aging monitoring is as follows: Basic information entry and binding: During step 2 initialization, attach an RFID tag to the drug and bring the portable RFID device close to the RFID tag. A prompt will pop up to update the drug's expiration date and confirm the product name. By mapping and binding the RFID tag's coding information with the item's name and expiration date, drug information is automatically uploaded to the hospital's emergency vehicle material management platform. Set the expiration date: The hospital emergency vehicle material management platform sets the corresponding expiration date according to the name and specifications of different drugs; Real-time monitoring and comparison: Continuously monitor the expiration date of drugs. The hospital emergency vehicle material management platform compares the remaining expiration date of the drug with the pre-set expiration date. Early warning: When the hospital emergency vehicle material management platform detects that the expiration date of the drug material enters the set near-expiration range, it will remind the user to dispose of the expiring drugs as soon as possible through voice and interface flashing, and turn off the alarm message after the user has handled it.

5. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In step 5, the doctor's medication order is converted into medication information in real time through the AI ​​voice recognition system, and the following process is specifically performed: Using dual-microphone beamforming technology combined with the RNN noise suppression model, the doctor's voice source is spatially located; An improved end-to-end recognition model based on the open-source Whisper architecture introduces a medical-specific vocabulary enhancement module, a context-aware dosage unit error correction mechanism, and real-time voiceprint verification to ensure the legitimacy of the medical order source; The conversion results are broadcast instantly through the TTS engine, and the doctor triggers the instruction status change through the preset command; Regular expressions and clinical terminology are combined for analysis and automatic conversion into standardized medical instructions.

6. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In step 6, the electrocardiogram image is captured by a portable terminal to automatically extract the vital signs of the patient. The specific identification process is as follows: Image preprocessing: Call the PyWT library wavelet denoising interface to perform denoising correction on the captured ECG image, and call the OpenCV perspective correction interface to automatically detect the vertices of the ECG image quadrilateral and map them to a standard rectangle; Vital sign data recognition: Using the open source model convolutional neural network (CNN) and the MIT-BIH dataset, we can classify heartbeats, including N, V, L, and R arrhythmias. At the same time, the open source model LightX3ECG was used for three-lead disease classification; Medical rule engine: Verify the consistency between model output and physiological logic, including: Conflict detection: If HR in the text area is 80 bpm, but the heart rate calculated by the RR interval in the waveform area is ≠ 80 ± 5%, a recheck will be prompted; Risk value interception: If ST segment elevation is detected and QTc>500ms, the portable terminal system will automatically warn of the risk of myocardial infarction and sound an alarm.

7. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 6 is characterized in that: In step five, the AI ​​speech recognition system has a directional noise suppression module, which uses dual-microphone beamforming technology combined with an RNN noise suppression model to spatially locate the doctor's voice source, and the software layer uses the MFCC-GMM acoustic model of the open source Kaldi framework to achieve environmental noise attenuation ≥15dB.

8. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In the step 1, the ledger established is the standard quantity for each material inventory, and the RFID electronic tag is pasted in a position that does not block the key information of the drug and has anti-falling properties.

9. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In the step 2, the coding rules include a hierarchical combination of region, hospital, department, emergency vehicle number and drug type.

10. The mobile emergency intelligent management method of RFID automatic identification and AI artificial intelligence interaction according to claim 1 is characterized in that: In step six, the extracted vital signs data include patient ID, time, lead name, heart rate value, waveform, and key diagnostic parameters, wherein the key diagnostic parameters include PR interval, QRS duration, and QT / QTc.

Images